High-Flow Nasal Cannulae in the Management of Apnea of Prematurity:
A Comparison With Conventional Nasal Continuous Positive
Airway Pressure
Con Sreenan, MB, MRCP(Ire)*‡; Robert P. Lemke, MSc, MD, FAAP, FRCPC*‡;
Ann Hudson-Mason, BSc, RRT*; and Horacio Osiovich, MD, FRCPC*‡
ABSTRACT. Apnea of prematurity (AOP) is frequently
managed with nasal continuous positive airway pressure
(NCPAP). Nasal cannula (NC) are used at low flows (<0.5
L/min) to deliver supplemental oxygen to neonates. A
number of centers use high-flow nasal cannula (HFNC)
in the management of AOP without measuring the pos-
itive distending pressure (PDP) generated.
Objective. To determine the NC flow required to gen-
erate PDP equal to that provided by NCPAP at 6 cm H2O
and to assess the effectiveness of HFNC as compared
NCPAP in the management of AOP.
Method. Forty premature infants, gestation 28.7 ⴞ 0.4
weeks (mean ⴞ standard error of mean), postconceptual
age at study 30.3 ⴞ 0.6 weeks, birth weight 1256 ⴞ 66 g,
study weight 1260 ⴞ 63 g who were being managed with
conventional NCPAP for at least 24 hours for clinically
significant apnea of prematurity, were enrolled in a trial
of ventilator-generated conventional NCPAP versus in-
fant NC at flows of up to 2.5 L/min. End expiratory
esophageal pressure was measured on NCPAP and on
NC, and the gas flow on NC was adjusted to generate an
end expiratory esophageal pressure equal to that mea-
sured on NCPAP. Two 6-hour periods were continuously
recorded and the data were stored on computer.
Results. The flow required to generate a comparable
PDP with NC varied with the infant’s weight and was
represented by the equation: flow (L/min) ⴝ 0.92 ⴙ 0.68x,
x ⴝ weight in kg, R ⴝ 0.72. There was no difference in the
frequency and duration of apnea, bradycardia or desatu-
ration per recording between the 2 systems.
Conclusion. NC at flows of 1 to 2.5 L/min can deliver
PDP in premature neonates. HFNC is as effective as
NCPAP in the management of AOP. Pediatrics 2001;107:
1081–1083; nasal cannula, apnea of prematurity, positive
distending pressure, esophageal pressure, nasal continuous
positive airway pressure.
ABBREVIATIONS. AOP, apnea of prematurity; NCPAP, nasal
continuous positive airway pressure; NC, nasal cannula; PDP,
positive distending pressure; HFNC, high-flow nasal cannula;
From the *Neonatal Intensive Care Unit, Royal Alexandra Hospital; and the
‡Department of Pediatrics, University of Alberta, Edmonton, Alberta, Can-
ada.
This work was presented in abstract form at the American Thoracic Society
International Conference; May 5–10, 2000; Toronto, Ontario; and the Society
for Pediatric Research meeting; May 12–16, 2000; Boston, MA.
Received for publication May 31, 2000; accepted Aug 25, 2000.
Reprint requests to (H.O.) Division of Neonatology, British Columbia Chil-
dren’s Hospital, 4480 Oak St, Vancouver, BC, Canada, V6H 3V4. E-mail:
hosiovich@cw.bc.ca
PEDIATRICS (ISSN 0031 4005). Copyright © 2001 by the American Acad-
emy of Pediatrics.
PEDIATRICS Vol. 107 No. 5 May 2001
Downloaded from pediatrics.aappublications.org at UCSF Kalmanovitz Library & CKM on December 7, 2014
NICU, neonatal intensive care unit; CPAP, continuous positive
airway pressure; EP, esophageal pressure.
A
pnea of prematurity (AOP) is frequently
managed with nasal continuous positive air-
way pressure (NCPAP).1–3 However, there
are a number of problems associated with its use.
Pressure effects can occur, which may lead to local
tissue necrosis with resulting nasal stenosis and de-
formity on healing.4,5 The prongs are irritating to the
nares and can increase nasal secretions and lead to an
increased risk of nasal infection.6 In our experience,
infants frequently become agitated to such a degree
that sedation may be required to maintain the prongs
in the nares.
Nasal cannula (NC) are used at low flows (⬍ 0.5L/
min) to deliver supplemental oxygen to neonates.
More recently it has been shown that NC can deliver
positive distending pressure (PDP) to premature ne-
onates if the flow is increased to 1 to 2 L/min (high-
flow nasal cannula [HFNC]).7 The pressure gener-
ated is determined by a number of factors including
the structure of the NC, gas flow through it, and the
anatomy of the infant’s airway. As most factors are
not variable, the PDP produced is directly propor-
tional to the gas flow rate.7
The aim of this study was to quantify the NC flow
required to generate PDP in premature neonates, and
to compare HFNC with NCPAP in the management
of apnea of prematurity. We hypothesized that the
flow required to generate PDP with HFNC would be
related to body weight, and that HFNC would not be
⬎20% less effective than NCPAP in reducing the
number and severity of apneas, desaturations, and
bradycardias in premature newborns with apnea
and bradycardia.
METHODS
The study was conducted at the neonatal intensive care unit
(NICU) at the Royal Alexandra Hospital between October 1998
and November 1999. All neonates admitted to our NICU with a
weight ⬍2.0 kg at the time of the study who had been receiving
NCPAP for at least 24 hours for AOP were eligible for the study.
All infants were receiving theophylline and had therapeutic levels
(55–110 ␮mol/L). In our unit, infants with AOP are initially com-
menced on theophylline. If they continue to have clinically signif-
icant apnea despite therapeutic theophylline levels, NCPAP is
started. Exclusion criteria included any congenital or chromo-
somal abnormalities, severe neurologic insults or neuromuscular
disease, and infants with active infection defined as a positive
blood or cerebrospinal fluid culture within the previous 48 hours.
1081
This study was reviewed and approved by the Health Research TABLE 1. Demographic Data for the 40 Infants Enrolled
Ethics Board of Faculty of Medicine at the University of Alberta.
Mean ⫾ Standard
Deviation (Range)
Design
The study had a crossover design which had within participant Gestational age (wk) 28.7 ⫾ 2.5 (24–33)
comparison. After informed parental consent, each infant started Postconceptual age at time of 30.3 ⫾ 3.8 (26.5–34)
the study with NCPAP (Argyle nasal CPAP cannula, Sherwood study (wk)
Medical, St Louis, MO). NCPAP was generated by the Infant Star Birth weight (g) 1256 ⫾ 417 (560–1950)
500 and 950 ventilators (Infrasonics Inc, San Diego, CA.) and set at Study weight (g) 1260 ⫾ 398 (660–2130)
6 cm H2O. NC set-up consisted of gas source, air-oxygen blender,
and a Hudson nonheated bubble humidifier (Hudson RCI, Te-
mecula, CA). After 6 hours, the infant was changed to infant NC H2O and with NC (4.65 ⫾ 0.02 cm H2O vs 4.53 ⫾ 0.02
(Salter Labs, Arvin, CA). Each infant was studied for 2 consecutive cm H2O, P ⫽ .84). As expected, the NC flow required
6-hour periods. The only change in the infant’s management was
the use of NC during the study. The infants were nursed in their to generate a PDP equal to that produced with NC-
isolettes in a thermoneutral environment, and feedings and care PAP increased with increasing infant weight (Fig 1)
were continued as previously. All infants were fed continuously and can be represented by the equation: flow (L/
via an orogastric tube. Oxygen saturations were maintained be- min) ⫽ 0.92 ⫹ 0.68 times the weight in kg (R ⫽ 0.72,
tween 88% and 94% for infants ⬍35 weeks and between 90% and
96% for those 35 weeks or more and those with bronchopulmo-
P ⬍ .001). We noted no significant difference in the
nary dysplasia. During the 12 hours of the study, oxygen satura- frequency and duration of apnea, bradycardia, or
tions, heart rate, respiratory rate, and apneas were monitored desaturation per recording between epochs using
continuously and recorded on computer (Asyst 401, Keithley In- NCPAP and NC (Table 2). Although on initial anal-
struments, Cleveland, OH). Any changes in Fio2 were recorded on ysis the mean duration of desaturations was shorter
a respiratory flow sheet.
with HFNC, this was no longer significant after
Measurement of Esophageal Pressure posthoc analysis. There were no side effects of NC
use noted (no difference in oxygen requirements, no
To assess the actual CPAP delivered for the level at which the
ventilator was set at that time, esophageal pressure (EP) was mucosal drying or trauma to the nares). No adverse
measured, as an indication of airway end-distending pressure, effect on feeding tolerance was noted with NC use.
with a saline-filled catheter (5-Fr Argyll, Sherwood Medical, St
Louis, MO) placed in the distal esophagus and attached to a DISCUSSION
differential pressure transducer (Cobe Labs, Lakewood, CO).7,8
The catheter was introduced into the stomach and then with-
This study demonstrates that ordinary NC can
drawn into the distal esophagus and positioned to achieve a wave deliver PDP at flows of up to 2.5 L/min in neonates
form that was free of cardiac artifact, negative during inspiration, up to 2.0 kg. In addition, we showed that NC are as
and flat during occlusion. The end-expiratory EP was defined as effective as NCPAP in the management of AOP with
the difference between EP at end-expiration and at baseline. When no difference in the number of apneas, bradycardias,
the CPAP was discontinued and before NC were placed, the EP
was documented to return to baseline. The infant was then placed or desaturation during a 6-hour period. Most impor-
on NC and the flow was adjusted to create equal CPAP to match tantly, oxygen requirements were not increased with
that delivered by the NCPAP. The esophageal catheter was kept in HFNC, and no mucosal drying or trauma to the
situ for 10 minutes to obtain stable readings, and it was then nares was noted.
removed and the flow kept constant during the study.
Using NC, the PDP delivered is determined by the
interaction of the NC, gas flow rate, and the anatomy
Statistical Analysis
of the infant’s airway.7 As expected, PDP increased
At the end of the study, the recordings were analyzed and the
number and duration of apneas, desaturations, and bradycardias
with increasing NC flow rates. Therefore, it is theo-
were documented while the infant was receiving either NCPAP or retically possible that using higher flows could gen-
NC. A significant apnea was defined as a cessation in breathing erate greater PDP. Conversely, in infants ⬎2.0 kg,
lasting at least 10 seconds associated with bradycardia and de- larger NC and higher flows would be needed to
saturation. Desaturation was defined as an oxygen saturation generate sufficient PDP. Although we did not find
⬍88% and bradycardia was defined as a drop in heart rate to
⬍70% of the baseline heart rate. Based on the previously pub-
lished variability of apnea in infants in our NICU,9 we calculated
that 40 infants would be required to show a 20% difference be-
tween the 2 groups, with a power of 0.8 and an ␣ of 0.05. Data are
expressed as either mean ⫾ standard error of the mean or stan-
dard deviation where appropriate. The data were analyzed using
analysis of variance. Posthoc analysis used Fisher’s least signifi-
cant difference test. The relationship between the flow (dependent
variable) required to generate an equal PDP to that produced by
NCPAP and birth weight (independent variable) was studied by
linear regression (Sigma Stat, Version 2.0, Jandel Corp, San Rafael,
CA). A P ⬍ .05 was considered significant.

RESULTS
Forty preterm infants were enrolled in the study,
and the demographic data are shown in Table 1.
Parental consent was denied in 2 cases and 2 infants
were excluded (1 with grade IV intraventricular
hemorrhage, and 1 with congenital myotonic dystro-
phy). As planned by the study design, there was no Fig 1. NC flow required to generate positive distending pressure
difference between the EP during NCPAP at 6 cm in preterm neonates.

1082 HIGH-FLOW NASAL CANNULA AND APNEA OF PREMATURITY

Downloaded from pediatrics.aappublications.org at UCSF Kalmanovitz Library & CKM on December 7, 2014
TABLE 2. Comparison Between HFNC and NCPAP
Parameter NCPAP NFNC
Apnea
Number (per 6 h) 1.6 ⫾ 0.5 2.0 ⫾ 0.8
Average duration (s) 22.5 ⫾ 2.0 24.4 ⫾ 1.9
Longest (s) 29.1 ⫾ 2.6 36.7 ⫾ 3.9
Bradycardias
Number (per 6 h) 2.6 ⫾ 0.5 3.5 ⫾ 0.8
Average duration (s) 26.3 ⫾ 2.0 24.5 ⫾ 2.0
Longest (s) 38.0 ⫾ 2.8 38.5 ⫾ 3.4
Lowest heart rate 80.1 ⫾ 2.2 78.1 ⫾ 2.3
Desaturations
Number (per 6 h) 7.5 ⫾ 1.6 6.2 ⫾ 1.3
Average duration (s) 32.0 ⫾ 1.8 25.7 ⫾ 1.8
Lowest saturation (%) 63.3 ⫾ 2.6 63.9 ⫾ 2.8
Fraction of inspired oxygen (%) 22.0 ⫾ 0.3 22.3 ⫾ 0.6
* Analysis of variance; data expressed as mean ⫾ standard error.
any undue drying effect on the nares from use of NC
at such flow rates, it is very likely that higher flow
rates could have a drying effect on the nares, which
would be minimized by incorporating a heated hu-
midifier into the system.
Our study confirms the findings of Locke et al7
who demonstrated that NC flow could deliver PDP
to infants and significantly alter breathing patterns.
However, they cautioned against the use of NC at
high flow in premature neonates, as uncontrolled
pressure may be delivered. In contrast, in our study
we measured EP as a measure of PDP and adjusted
the NC flow to provide the same EP as provided by
the level of NCPAP that we use in clinical practice.
We did not measure pharyngeal pressure in addition
to esophageal pressure because pharyngeal stimula-
tion in itself may influence the occurrence of apnea.10
The choice of a 6-hour before and after crossover
was arbitrary. This time period was chosen as it has
been previously shown in our NICU and by other
centers to be long enough to document a minimum
number of apnea episodes.9,11,12 Based on a sample
size of 40 infants, we can be confident that a 20% or
greater increase in the incidence of apnea during
HFNC was not missed by our study. After crossover
between NCPAP and HFNC, it is possible that there
is a residual effect of treatment that the study time of
6-hours may have concealed. Given the large sample
size, we think it is unlikely that such an effect was
missed. Theophylline dose had not been adjusted
within 3 days of the study in any infant, and as such,
should have been at a steady state. Serum theophyl-
line levels were therapeutic (range: 55–110 ␮mol/L)
and are unlikely to have fluctuated widely during
the study period.
Although HFNC was as effective as NCPAP with
respect to the number of central apneas, we were
unable to document changes in obstructive and
mixed apneas. However, as both methods provided
equal distending pressure in the upper airway, pas-
sive splinting should be the same.3 In addition, an
increase in obstructive apneas with HFNC would
have resulted in an increased number of desatura-
Downloaded from pediatrics.aappublications.org at UCSF Kalmanovitz Library & CKM on December 7, 2014
tions. However, this did not occur. As we did not
P* perform any measurements of pulmonary mechanics
in general and work of breathing in particular, we
.7
cannot yet recommend the use of HFNC as a method
.67 of respiratory support for lung disease in the neo-
.31 nate. The NC set-up used in this study used a non-
heated humidifier. Although we did not note any
.33 mucosal drying effect with NC use, the study time
.62
.93 was only 6 hours. For longer use in our NICU we
.63 have now incorporated a heated humidifier into the
NC set-up.
.52 A potential drawback of HFNC is that the pressure
.04
.89
generated by the airflow is dependent on mainte-
.71 nance of a good seal in the oral cavity. If the infant
mouth breaths then airflow escapes through the
mouth, and distending pressure may be lost. This
also occurs to a lesser degree during NCPAP. Mouth
breathing may explain why HFNC may not be an
effective treatment in some infants with AOP. To
minimize some of these problems, infants were
maintained in the same position throughout the
study.
CONCLUSION
At flows of up to 2.5 L/min in infants ⬍2.0 kg,
HFNC can generate PDP which is as effective as
NCPAP in the management of AOP. It is easy to
perform and is well-tolerated. We, therefore, recom-
mend HFNC as a way of providing PDP in infants
with AOP.
REFERENCES
1. Milner AD. Apnea and bradycardia. In: Rennie JM, Roberton NRC, eds.
Textbook of Neonatology. 3rd ed. Edinburgh, Scotland: Churchill
Livingstone; 1999:630 – 637
2. Kattwinkel J, Nearmann HS, Fanaroff AA, Katona PG, Klaus MH.
Apnea of prematurity: comparative effects of cutaneous stimulation and
nasal CPAP. J Pediatr. 1975;86:588 –592
3. Martin RJ, Nearman HS, Katona PG, Klaus MH. The effect of a low
continuous positive airway pressure on the reflex control of respiration
in the preterm infant. J Pediatr. 1977;90:976 –981
4. Robertson NJ, Mc Carthy LS, Hamilton PA, Moss AL. Nasal deformities
resulting from flow driver continuous positive airway pressure. Arch
Dis Child. 1996;75:F209 –F212
5. Ahluwalia JS, White DK, Morley CJ. Infant flow driver or single prong
nasal continuous positive airway pressure: short-term physiological
effects. Acta Paediatr. 1998;87:325–327
6. Ahmuda CA, Goldsmith JP. Continuous distending pressure. In: Gold-
smith JP, Karotkin EH, eds. Assisted Ventilation of the Neonate. 3rd ed.
Philadelphia, PA: WB Saunders; 1996
7. Locke RG, Wolfson MR, Shaffer TH, Rubenstein D, Greenspan JS.
Inadvertent administration of positive end-expiratory pressure during
nasal cannula flow. Pediatrics. 1993;91:135–138
8. Coates AL, Davis GM, Vallinis P, Outerbridge EW. Liquid-filled esoph-
ageal catheter for measuring pleural pressure in preterm neonates.
J Appl Physiol. 1989;67:889 – 893
9. Ryan CA, Finer NN, Peters KL. Nasal intermittent positive-pressure
ventilation offers no advantage over nasal continuous positive airway
pressure in apnea of prematurity. Am J Dis Child. 1989;143:1196 –1198
10. Thach BT. Reflux associated apnea in infants: evidence for a laryngeal
chemoreflex. Am J Med. 1997;103:120 –124
11. Lin CH, Wang ST, Lin YJ, Yeh TF. Efficacy of nasal intermittent positive
pressure ventilation in treating apnea of prematurity. Pediatr Pulmonol.
1998;26:349 –353
12. Kurz H. Influence of nasopharyngeal CPAP on breathing pattern and
incidence of apneas in preterm infants. Biol Neonate. 1999;76:129 –133
ARTICLES 1083
High-Flow Nasal Cannulae in the Management of Apnea of Prematurity: A
Comparison With Conventional Nasal Continuous Positive Airway Pressure
Con Sreenan, Robert P. Lemke, Ann Hudson-Mason and Horacio Osiovich
Pediatrics 2001;107;1081
DOI: 10.1542/peds.107.5.1081
Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/107/5/1081.full.h
tml
References This article cites 10 articles, 2 of which can be accessed free
at:
http://pediatrics.aappublications.org/content/107/5/1081.full.h
tml#ref-list-1
Citations This article has been cited by 26 HighWire-hosted articles:
http://pediatrics.aappublications.org/content/107/5/1081.full.h
tml#related-urls
Subspecialty Collections This article, along with others on similar topics, appears in the
following collection(s):
Fetus/Newborn Infant
http://pediatrics.aappublications.org/cgi/collection/fetus:newb
orn_infant_sub
Permissions & Licensing Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
http://pediatrics.aappublications.org/site/misc/Permissions.xht
ml
Reprints Information about ordering reprints can be found online:
http://pediatrics.aappublications.org/site/misc/reprints.xhtml

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

publication, it has been published continuously since 1948. PEDIATRICS is owned, published,
and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk
Grove Village, Illinois, 60007. Copyright © 2001 by the American Academy of Pediatrics. All
rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at UCSF Kalmanovitz Library & CKM on December 7, 2014
High-Flow Nasal Cannulae in the Management of Apnea of Prematurity: A
Comparison With Conventional Nasal Continuous Positive Airway Pressure
Con Sreenan, Robert P. Lemke, Ann Hudson-Mason and Horacio Osiovich
Pediatrics 2001;107;1081
DOI: 10.1542/peds.107.5.1081

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/107/5/1081.full.html

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

publication, it has been published continuously since 1948. PEDIATRICS is owned,
published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point
Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2001 by the American Academy
of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at UCSF Kalmanovitz Library & CKM on December 7, 2014